In a continued coordinated effort, using the techniques of theoretical chemistry, synthetic chemistry, receptor-binding and animal testing, two major objectives will be addressed: (1) Identification and characterization of molecular properties leading to each of the four benzodiazepine-receptor modulated activities and antagonism to them and the relationship of these in vivo activities to possible multiple BZ-receptor types. (2) Design of new analogs with specific combinations of agonist and antagonist activities which could be anxioselective agents of low abuse potential. To accomplish the first major objective, common molecular requirements for each agonist activity will be determined by comparisons of calculated conformational and electronic properties of benzodiazepines and three more specific classes of anxioloytic agents. In a similar fashion, common molecular requirements for antagonism to each activity in three types of potent benzodiazepine antagonists with varying selectivity will be investigated. In a parallel effort, studies to clarify the question of BZ receptor heterogeneity will continue using computer- assisted analysis of receptor-binding experiments conducted with a variety of selective and nonselective agonists and antagonists that complete for the BZ receptor site(s). These theortical and experimental studies together should lead to the understanding of requirements for selective agonism and antogonism. In a second major effort, molecular indicators of agonism and antagonism will be used to design and synthesize candidate high- affinity, selective agonist/antagonists. Receptor-binding and animal-testing studies will be performed to test these hypotheses. The in vivo potency of these compounds will be assessed for the four major activities of the BZs, i.e., anxiolytic, anticonvulsant, sedative-hypnotic, and muscle-relaxant properties. Possible antagonism to these actions will also be determined. Analogs that display selectivity or a combination of agonist and antagonist activity will be monitored for potentiation of the CNS depressant effects of ethanol and barbiturates. Physical-dependence liability will be assessed by determining whether antagonists can precipitate withdrawal symptoms. It is possible that among these new analogs will be anxioselective agents of low abuse potential that will not potentiate the CNS depressant effects of ethanol and barbiturates.